What is DBS's AI strategy?
DBS's stated aim, in CEO Tan Su Shan's words, is to be 'an AI-enabled bank with a heart'. The bank has invested in AI for more than a decade, and Tan told Euromoney in July 2026 that since its digitalisation journey began in 2014, 'our early investments in digital and AI laid a strong foundation'. The 2025 annual report counts more than 2,000 models across more than 430 use cases and approximately SGD 1 billion of economic value in 2025. In May 2026 Tan told The Business Times that agentic AI is the next frontier, seen through 'three lenses – personal agents, team agents and enterprise agents', with 10,000 personal agents built since late 2025 and enterprise work under way in 11 areas. On 6 August 2026 she said DBS has no 2026 target for AI economic value, and told CNA the same day that AI was beginning to contribute to fee income growth. Technology spending is about 10% of revenue and roughly flat, per Tan in August 2026; DBS does not disclose an AI budget. On 12 October 2026 it held its first Asia Future Forward Conference, on the theme of 'Intentional AI'.[1][2][3][4][5][6][7]
Data as of Nov 17, 2026 · DBS profile updated Nov 15, 2026 · 27 sources cited on this page · 141 banks profiled on this site · Markdown
What has DBS's leadership said about AI, and when?
In date order. Words in quotation marks are as published by the cited source.
| Date | Who | What was said | Where |
|---|---|---|---|
| November 6, 2025 | Tan Su Shan, CEO | The bank was on track to see S$1 billion in incremental impact from AI that year, from about 1,500 AI models across more than 370 use cases; '(AI) is literally embedded in our processes already' | CNA[8] |
| November 14, 2025 | Tan Su Shan | 'It's not hope. It's now. It's already happening. And it will get even better,' | CNBC[9] |
| February 9, 2026 | Tan Su Shan | More than 60% of staff were very actively using DBS-GPT; AI moving from efficiency tool to growth engine | Q4 2025 briefing (The Business Times)[10] |
| May 19, 2026 | Tan Su Shan | The leadership offsite of 200 to 300 senior leaders wore T-shirts reading 'Talk is cheap. Show me your agent.'; agentic AI is the next frontier, seen through personal, team and enterprise agents | The Business Times[1] |
| July 17, 2026 | Tan Su Shan | 'Last year, we scaled up our agentic AI deployment across the bank – which we see as the next wave of growth' | Euromoney[3] |
| August 6, 2026 | Tan Su Shan | Asked about a 2026 AI value target: 'We do not. It is difficult to measure because there are different types of AI.' | Q2 2026 media briefing transcript[5] |
| August 6, 2026 | Tan Su Shan | 'We're beginning to see the green shoots of AI working in generating ideation that generates fees' | CNA[6] |
| August 7, 2026 | Tan Su Shan | Total technology spending is about 10% of revenue and flat; AI token costs per use are falling | Bloomberg Television via Yahoo Finance[7] |
| September 27, 2026 | Tan Su Shan | DBS seeks to be a trusted and neutral interlocutor that connects capital, fosters innovation and provides trusted governance for AI in Asia; conference theme 'Intentional AI' | The Business Times[2] |
| October 12, 2026 | Tan Su Shan | On adding AI to old workflows: 'We call that AI lipstick. That's not going to work.' | The Straits Times[11] |
How did DBS's AI strategy develop?
DBS's first AI experiment did not work. In 2014 it tried DBS IBM Watson to process financial data and generate wealth-management advice; Tan told The Business Times it achieved only about 80 per cent accuracy, and she cites it as the reason the bank accepts 'that trough of disappointment'.[1]
The bank then built its data base first. Tan told CNA in November 2025 that DBS 'started many years ago to create our own data lake, set up processes and protocols, data security, and philosophy around how we deal with data', which she credits for its head start.[8]
The current phase is agentic. Tan described the agent registry and control plane in May 2026: 'We can see all the agents, the agent registry, with numbers and identification details,' with accountability, observability, traceability and evaluations of the agents. She said the business payoff is now showing in fees, with AI helping relationship managers generate ideas and 'nudging customers the right way'.[1][6]
What is not disclosed: DBS publishes no AI budget, no 2026 target for AI economic value and no figure for the value of generative or agentic AI separately from the classic-AI figure of about SGD 1 billion for 2025.[5][7][4]
What does the DBS profile record on this topic?
The rows below are the entries of the DBS Bank AI strategy profile that bear on this page, as of Nov 15, 2026; the profile holds the full record of 34 dated moves, 17 use cases and 51 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| Agentic credit-memo chain for corporate lending | A chain of roughly 70 to 80 agents assembles credit memos for large corporate clients from annual reports, news and internal notes; the relationship manager interrogates and finalises the draft, and DBS caps the steps in any chain with checks between stages. On 19 August 2026 DBS said the solution had been rolled out to approximately 1,500 employees globally after a 150-person pilot. | Rolling out | [12][13][14][15] |
| Agentic commerce with Visa Intelligent Commerce | Pilot of AI agents paying on behalf of consumers with tokenised DBS/POSB card credentials under issuer-controlled authentication. | Pilot | [16][14] |
| Multi-model sourcing and token-cost control | The CSO Assistant runs on Google's Gemini and Anthropic's Claude, models are reached through APIs on platforms such as Vertex AI, and DBS runs proofs of concept to see which provider is best at what. A memory cache, several providers and routing simple queries to small language models cut the unit cost of AI use. | In production | [17][18][7] |
| Responsible Data Use framework and on-premises DBS-GPT | Every AI use case is checked against a data foundation (security, privacy, access, quality), the PURE test and model governance. DBS-GPT sits on the bank's own premises and does not let APIs pull all its data; Tan stresses data lifecycle management from creation to deletion. | In production | [19][20] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Oct 13, 2026 | Global Finance: World's Best Bank for AI in Finance (Corporate) again, plus best AI in fraud prevention | Fintech News Singapore reports that at Global Finance's awards on 13 October 2026 DBS was named World's Best Digital Bank, the first Asia-headquartered winner in the programme's 27 years, and that, according to the bank, it won World's Best Bank for AI in Finance - Corporate for the second year running and the global Best AI in Fraud Detection/Prevention award in the corporate and institutional category; Tan Su Shan was quoted on the 'ambition of being an AI-enabled Bank with a Heart'. This is third-party recognition, not a new deployment. | [21] |
| Oct 12, 2026 | Tan at DBS Asia Future Forward: 'AI lipstick' and SGD 25 billion of AI financing | Tan told the inaugural conference that adding AI to old workflows will not change how work is done, and told Bloomberg Television that DBS had originated about SGD 25 billion of AI deals in 2026, mostly hardware, holding only a fraction. | [11][22] |
| Aug 19, 2026 | Tan: AI pace may force a resiliency versus speed trade-off | At the Singapore Computer Society Tech3 Forum Tan said staff agents are built on DBS-GPT, which sits on the bank's own premises and does not let APIs pull all its data, and that agents touching production face tighter limits and the full software lifecycle. | [20] |
| Aug 7, 2026 | Tan on the AI token paradox and technology spend | Tan told Bloomberg Television that DBS consumes far more AI tokens at a lower unit cost through caching, several providers and small language models, that total technology spending is about 10% of revenue and flat, and that the bank runs proofs of concept across Western and Chinese models. | [7] |
| Aug 6, 2026 | Q2 briefing: no 2026 AI value target, model-agnostic stack | Tan said DBS has no 2026 target for AI economic value, is focusing agentic work on 12 enterprise journeys, and uses 'US, Chinese, everything' while avoiding reliance on any one provider. | [5] |
| May 19, 2026 | 'Show me your agent': Tan outlines personal, team and enterprise agents | Tan told The Business Times that staff had created 10,000 personal agents since late 2025, that seven in 10 employees use Gen AI tools (1.8 million prompts a month), and that enterprise-agent work was under way in 11 areas. | [13] |
| Apr 30, 2026 | Tan on Anthropic Mythos: it amplifies cyber risk | After first-quarter results Tan said Mythos could help attackers find weaknesses faster and at greater volume but could also be used by banks to defend; the immediate task was patches and strong internal hygiene, and DBS still saw AI as a net positive. MAS had convened bank CEOs on AI cyber threats. | [23][24] |
| Mar 9, 2026 | 2025 annual report: 2,000 models, 430 use cases, about SGD 1 billion | Reports approximately SGD 1 billion of economic value for 2025, DBS-GPT open to all employees in core markets, nine operating-model transformations completed against a target of six, and a move 'from AI as a copilot to AI operating on autopilot' under governance. | [4] |
| Feb 9, 2026 | Q4 2025 briefing: DBS-GPT use above 60%, AI value to become harder to isolate | Tan said more than 60% of staff were very actively using DBS-GPT, which was not so good when launched in July 2025, credited AI nudges for new-to-bank customers and deposit inflows, and said DBS might keep A/B testing but expected more value to come as capacity built for growth; AI had compressed technical-debt work from months or years to weeks. | [10] |
| Nov 14, 2025 | Tan: AI revenue bump above SGD 1 billion in 2025 | Tan told CNBC at Singapore Fintech Week that AI adoption was already paying off, based on about 370 use cases powered by over 1,500 models, against SGD 750 million in 2024. | [9] |
| Nov 7, 2025 | DBS stops hiring for roles AI is expected to replace | Tan told Bloomberg Television the bank would not hire for jobs that will be replaced by AI and is retraining affected staff, for example tellers moving to video teller machines or relationship-manager roles. | [25] |
| Jun 19, 2025 | Head of generative AI: 1,500 models and over 350 use cases | Rajeev Hassamal, DBS head of generative AI and flow of work, told SuperAI Singapore that DBS had more than 350 use cases and 1,500 models in production, with AI earning about SGD 750 million and an economic value above SGD 1 billion expected in 2025, and showed DBS-GPT. | [26][27] |
What is DBS's AI strategy in one line?
To be 'an AI-enabled bank with a heart', in CEO Tan Su Shan's words: AI embedded across the bank, now extended to personal, team and enterprise agents, with governance and staff retraining alongside.
How much value does DBS say AI creates?
Approximately SGD 1 billion of economic value in 2025 from more than 2,000 models across more than 430 use cases, per the 2025 annual report. Tan said in August 2026 there is no 2026 target.
How much does DBS spend on AI?
DBS does not disclose AI spending. Tan told Bloomberg Television in August 2026 that total technology spending is about 10% of revenue and roughly flat, and that the cost of each AI query is falling as the bank uses caching, several providers and small language models.
Is AI adding revenue at DBS?
Tan told CNA on 6 August 2026 that AI was beginning to contribute to fee income growth in wealth management and corporate banking, calling it 'the green shoots of AI working in generating ideation that generates fees'.
When did DBS start with AI?
DBS has invested in AI for more than a decade. In 2014 it piloted DBS IBM Watson for wealth advice, which reached only about 80 per cent accuracy, before building its data platforms.
- From 'show me your code' to 'show me your agent': DBS CEO Tan Su Shan on how agentic AI is transforming the bank
- Discussing the future of AI in Singapore and Asia: DBS conference
- Asia's best digital bank 2026: DBS
- Annual Report 2025
- Edited transcript of DBS second-quarter 2026 media briefing, 6 August 2026
- DBS CEO says AI is beginning to contribute to fee income growth
- DBS CEO sees AI costs falling with more use in 'token paradox'
- DBS leans on AI efficiency as interest rate pressures squeeze margins
- CEO of Southeast Asia's largest bank says AI adoption already paying off: 'It's not hope, it's now'
- DBS says AI push is paying off as it shifts from efficiency tool to growth engine
- Human judgment and intentional outcomes are vital in AI use: DBS, TikTok CEOs say
- DBS holds off on letting AI agents run on their own as controls lag capability (undated page, retrieved 2026-10-05)
- CEO Tan Su Shan on how agentic AI is transforming the bank
- DBS Group CEO Tan Su Shan Leads Southeast Asia's Largest Bank Into The AI Era
- DBS scales agentic AI to transform way of working for corporate bankers, freeing up time for more strategic client engagements
- DBS is first bank in Asia Pacific to pilot Visa Intelligent Commerce for everyday payments
- CW Innovation Awards: DBS AI assistant slashes call times and boosts productivity (undated page, retrieved 2026-10-14)
- How Singapore's largest bank builds AI with confidence (interview with Nimish Panchmatia)
- Ethical & Responsible AI in Banking
- DBS CEO says AI's pace could test corporate safeguards
- DBS Named World’s Best Digital Bank 2026 by Global Finance
- 'The market has spoken': DBS CEO Tan Su Shan says market is repricing AI financing risks
- Banks in Asia brace for complex cyber threats from frontier AI
- Wealth ambitions, AI threats and war risks: 4 takeaways from Singapore banks' Q1 results
- Singapore's DBS girds for AI as tellers morph into bankers
- DBS expects economic value from its AI use to top $1 billion in 2025
- DBS expects economic value from its use of AI to exceed S$1 billion in 2025
When DBS moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning